Doctor of Philosophy (PhD)
Faculty of Education
Dalgarno, Barney J, Characteristics of 3D environments and potential contributions to spatial learning, PhD thesis, Faculty of Education, University of Wollongong, 2004. http://ro.uow.edu.au/theses/214
Three dimensional (3D) technologies have revolutionised computer games to the extent that virtually all new games are based upon 3D graphics. Some might claim that it is only a matter of time before 3D environments become the norm for other types of software, such as business systems, desktop computer task managers and online learning resources. On the surface it would appear that 3D environments have great potential in educational contexts as they provide the possibility of rich learner engagement and allow learners to explore, construct and manipulate virtual objects, structures and metaphorical representations of ideas. This thesis argues that the potentially unique contribution of such environments to learning depends on their ability to facilitate spatial learning. Specifically, it is argued that the ability of such environments to facilitate spatial learning is an implicit assumption in applications of 3D learning environments, whether they are based on models of real or metaphorical objects or spaces. In order to explore the degree to which 3D environments have unique advantages over alternatives such as video or interactive multimedia in facilitating spatial learning, it was necessary to first identify the distinguishing characteristics of 3D learning environments. The study identified seven characteristics of 3D learning environments that distinguish them from other types of multimedia learning resources. These are realistic display, smooth update of views, smooth display of object motion, consistency of object behaviour, control of view position and direction, object manipulation, and control of object model and simulation parameters. Having identified the distinguishing characteristics of 3D environments, the study explored the contribution of these characteristics to spatial learning. Investigating the contribution of all of these characteristics was considered to be outside the scope of a doctorate. Consequently, the contribution to spatial learning of smooth display of view changes, smooth display of object motion, user control of view position and direction, and object manipulation were investigated. Additionally, the study explored the effect of learning task design within a 3D environment on spatial learning. Versions of a 3D environment modelled on a chemistry laboratory were developed for use as research instruments, each with the inclusion or exclusion of some of the identified distinguishing characteristics. Participants in the study used these environments and then undertook tests to determine the degree to which they formed a spatial cognitive model of the laboratory and its apparatus. Quantitative data analysis techniques were used to compare their test performances. This allowed conclusions to be reached about the contribution of each of the identified characteristics to spatial learning. Three phases of investigation were carried out. The first, a pilot investigation, used qualitative methods to explore the usability of the virtual environment and to explore aspects of the learning that occurred through its use. In the first major investigation, three groups of participants were used. One group explored a version of the virtual laboratory with smooth view changes, smooth display of object motion, user control over view and object manipulation. A second group explored a version of the virtual laboratory without user control over view or object manipulation capability. A third group explored the real laboratory. In the second major investigation, three groups of participants were again used. One group again used a version of the virtual laboratory with all of the identified characteristics. A second group explored a version of the virtual laboratory without user control over view or object manipulation capability. A third group explored a version of the virtual laboratory without user control over view or object manipulation capability and without smooth display of view changes and smooth display of object motion. Additionally, the task carried out by the first group was varied in this second investigation in order to explore the contribution of learning task design to spatial learning. Smooth display of view changes was found to contribute to spatial learning in some but not all circumstances, and user control over view position and direction was found to contribute to spatial learning only when the task carried out in the environment was closely aligned with the desired learning. The results provided little support for the contribution of smooth display of object motion or object manipulation to spatial learning, but there were limitations in the complexity of the objects explored and the range of object manipulations carried out. The results have implications for educational designers considering the development or use of a 3D learning environment and needing to make a decision between this and alternatives such as static images and video. The advantages of such environments over video depend on the degree to which the environments allow tasks to be performed that directly align with the desired learning outcomes. If such tasks can be identified then learning advantages can occur, but only if learners are explicitly advised to undertake these tasks either through guidance provided within the environment or as part of supporting materials. The free exploration of a 3D environment with no explicit task advice is unlikely to lead to learning advantages over video or interactive multimedia.